[dpdk-dev] [PATCH v2] ethdev: add field for device data per process

Alejandro Lucero alejandro.lucero at netronome.com
Fri Oct 5 19:00:08 CEST 2018


On Fri, Oct 5, 2018 at 5:37 PM Andrew Rybchenko <arybchenko at solarflare.com>
wrote:

> On 10/5/18 7:16 PM, Alejandro Lucero wrote:
>
> Primary and secondary processes share a per-device private data. With
> current design it is not possible to have data per-device per-process.
> This is required for handling properly the CPP interface inside the NFP
> PMD with multiprocess support.
>
> There is also at least another PMD driver, tap, with similar
> requirements for per-process device data.
>
> v2:
>  - changing library version
>  - report shared library change in release notes
>
> Signed-off-by: Alejandro Lucero <alejandro.lucero at netronome.com> <alejandro.lucero at netronome.com>
> ---
>  doc/guides/rel_notes/release_18_11.rst | 1 +
>  lib/librte_ethdev/Makefile             | 2 +-
>  lib/librte_ethdev/meson.build          | 2 +-
>  lib/librte_ethdev/rte_ethdev_core.h    | 8 +++++++-
>  4 files changed, 10 insertions(+), 3 deletions(-)
>
> diff --git a/doc/guides/rel_notes/release_18_11.rst b/doc/guides/rel_notes/release_18_11.rst
> index c806dc6..32ab754 100644
> --- a/doc/guides/rel_notes/release_18_11.rst
> +++ b/doc/guides/rel_notes/release_18_11.rst
> @@ -189,6 +189,7 @@ Shared Library Versions
>       librte_acl.so.2
>     + librte_cfgfile.so.2
>       librte_cmdline.so.2
> +   + librte_ethdev.so.2
>
>
> It is a comment section. Actual library version is librte_ethdev.so.10
> below.
> So, it should add + sign and change to so.11
>
>
I has the + sign but you are right about the version number. I will fix
that.
Thanks

>
>     This section is a comment. Do not overwrite or remove it.
>     =========================================================
> diff --git a/lib/librte_ethdev/Makefile b/lib/librte_ethdev/Makefile
> index d720dd2..e27bcd5 100644
> --- a/lib/librte_ethdev/Makefile
> +++ b/lib/librte_ethdev/Makefile
> @@ -16,7 +16,7 @@ LDLIBS += -lrte_mbuf -lrte_kvargs
>
>  EXPORT_MAP := rte_ethdev_version.map
>
> -LIBABIVER := 10
> +LIBABIVER := 11
>
>  SRCS-y += ethdev_private.c
>  SRCS-y += rte_ethdev.c
> diff --git a/lib/librte_ethdev/meson.build b/lib/librte_ethdev/meson.build
> index 172e302..6783013 100644
> --- a/lib/librte_ethdev/meson.build
> +++ b/lib/librte_ethdev/meson.build
> @@ -2,7 +2,7 @@
>  # Copyright(c) 2017 Intel Corporation
>
>  name = 'ethdev'
> -version = 10
> +version = 11
>  allow_experimental_apis = true
>  sources = files('ethdev_private.c',
>  	'ethdev_profile.c',
> diff --git a/lib/librte_ethdev/rte_ethdev_core.h b/lib/librte_ethdev/rte_ethdev_core.h
> index 33d12b3..0d28fd9 100644
> --- a/lib/librte_ethdev/rte_ethdev_core.h
> +++ b/lib/librte_ethdev/rte_ethdev_core.h
> @@ -539,7 +539,13 @@ struct rte_eth_dev {
>  	eth_rx_burst_t rx_pkt_burst; /**< Pointer to PMD receive function. */
>  	eth_tx_burst_t tx_pkt_burst; /**< Pointer to PMD transmit function. */
>  	eth_tx_prep_t tx_pkt_prepare; /**< Pointer to PMD transmit prepare function. */
> -	struct rte_eth_dev_data *data;  /**< Pointer to device data */
> +	/**
> +	 * Next two fields are per-device data but *data is shared between
> +	 * primary and secondary processes and *process_private is per-process
> +	 * private. The second one is managed by PMDs if necessary.
> +	 */
>
>
> I think above is wrong from doxygen point of view. 'data' member already
> has
> documentation. I'm not sure, so it should be checked.
>
>
I though doxygen would use just the comment after the fields declaration.
Not sure how to handle this.
Should I remove the comment and just leave the next two lines?

> +	struct rte_eth_dev_data *data;  /**< Pointer to device data. */
> +	void *process_private; /**< Pointer to per-process device data. */
>  	const struct eth_dev_ops *dev_ops; /**< Functions exported by PMD */
>  	struct rte_device *device; /**< Backing device */
>  	struct rte_intr_handle *intr_handle; /**< Device interrupt handle */
>
>
>


More information about the dev mailing list